INDEX

Explanations

phrases related to technical instructions or procedures

oai_token-act-pair · gpt-3.5-turbo

sentences that convey completed actions or statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 prophes

-0.77

 cheering

-0.71

 damning

-0.70

 betrayed

-0.70

 accuser

-0.69

 abruptly

-0.68

plet

-0.68

 offended

-0.67

ŃĶ

-0.67

 unmist

-0.66

POSITIVE LOGITS

 Alternatively

1.64

 However

1.39

 Examples

1.36

 Typically

1.32

 Others

1.26

 Often

1.25

 Depending

1.25

 Generally

1.23

 Likewise

1.23

 Usually

1.21

Activations Density 0.546%

phrases related to technical instructions or procedures

sentences that convey completed actions or statements

No Comments

No Known Activations