INDEX

Explanations

verbs or phrases indicating decision-making or actions taken

oai_token-act-pair · gpt-3.5-turbo

verbs related to actions or decisions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

assic

-0.75

fortunately

-0.65

 evidenced

-0.64

 Glob

-0.61

itcher

-0.60

 fortunately

-0.59

stellar

-0.58

surv

-0.58

acs

-0.58

 wont

-0.58

POSITIVE LOGITS

enance

0.94

igate

0.85

dress

0.79

him

0.71

itate

0.71

uate

0.70

 them

0.70

 seiz

0.68

 withdrawals

0.68

escription

0.68

Activations Density 0.395%

verbs or phrases indicating decision-making or actions taken

verbs related to actions or decisions

No Comments

No Known Activations