INDEX

Explanations

mentions of approvals or agreements

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "approve" and its variations

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

ngth

-0.74

tone

-0.72

grave

-0.71

lla

-0.70

ottesville

-0.68

toe

-0.68

tal

-0.66

monds

-0.66

 Islands

-0.65

 Hole

-0.65

POSITIVE LOGITS

 unanimously

1.03

 approve

0.98

 approval

0.97

 approvals

0.96

 approving

0.92

 authorizing

0.92

 approves

0.91

votes

0.80

 ratings

0.79

clinton

0.74

Activations Density 0.029%

mentions of approvals or agreements

occurrences of the word "approve" and its variations

No Comments

No Known Activations