INDEX

Explanations

statements or actions that are repeated multiple times

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "repeatedly."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.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

Reviewer

-0.79

igans

-0.77

toc

-0.76

ocard

-0.73

ghai

-0.73

istan

-0.69

LCS

-0.69

ocene

-0.68

Soc

-0.66

oke

-0.66

POSITIVE LOGITS

theless

0.97

 unsuccessful

0.91

 reiter

0.89

 interrupted

0.87

 reaff

0.85

 reiterated

0.85

 repeated

0.84

 contradicted

0.83

 voiced

0.81

 stressed

0.80

Activations Density 0.053%

statements or actions that are repeated multiple times

occurrences of the word "repeatedly."

No Comments

No Known Activations