INDEX

Explanations

references to specific individuals or characters, potentially focusing on actions or descriptions related to them

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 specialization

-0.75

 promotions

-0.70

 Revenue

-0.69

 Comprehensive

-0.68

ontent

-0.68

Rum

-0.67

 Savings

-0.66

 rumors

-0.66

 Residents

-0.65

 specializing

-0.65

POSITIVE LOGITS

stretched

1.38

 fists

1.09

 clasp

1.07

 shoulder

1.07

 perpendicular

1.06

 crou

1.05

 grasped

1.05

 grasping

1.05

 gently

1.03

 crotch

1.03

Activations Density 2.292%

references to specific individuals or characters, potentially focusing on actions or descriptions related to them

No Comments

No Known Activations