INDEX

Explanations

references to specific brands or entities like "Cisco", "Db", "Football Reference", "Torah", "Amazon", "Huawei", "FCC", "Samsung", "Air Force", "Simulink", "HS", "FP", "AU", "Blender", and "Post

oai_token-act-pair · gpt-3.5-turbo

specific brand names and technical terms related to technology and product specifications

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Amid

-0.74

Amid

-0.70

 amid

-0.68

oken

-0.67

 amidst

-0.66

actionDate

-0.65

assis

-0.65

meet

-0.65

 Rise

-0.65

joining

-0.65

POSITIVE LOGITS

 recommends

1.81

 prefers

1.56

 advises

1.46

 calculates

1.42

 deems

1.38

 assigns

1.37

 uses

1.37

 warns

1.37

 considers

1.36

 forbids

1.36

Activations Density 0.499%

references to specific brands or entities like "Cisco", "Db", "Football Reference", "Torah", "Amazon", "Huawei", "FCC", "Samsung", "Air Force", "Simulink", "HS", "FP", "AU", "Blender", and "Post

specific brand names and technical terms related to technology and product specifications

No Comments

No Known Activations

references to specific brands or entities like "Cisco", "Db", "Football Reference", "Torah", "Amazon", "Huawei", "FCC", "Samsung", "Air Force", "Simulink", "HS", "FP", "AU", "Blender", and "Post

specific brand names and technical terms related to technology and product specifications

No Comments

No Known Activations