INDEX

Explanations

words related to negative evaluations or criticisms

oai_token-act-pair · gpt-3.5-turbo

contexts that describe inadequately executed actions or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

dar

-0.89

ership

-0.82

thood

-0.79

thouse

-0.78

hip

-0.74

rage

-0.73

itar

-0.72

vill

-0.72

fig

-0.72

aire

-0.70

POSITIVE LOGITS

nesses

0.75

 misc

0.74

 mishand

0.70

 botched

0.69

 underestimate

0.69

 miser

0.68

 inaccurate

0.68

 undermin

0.67

 wasteful

0.67

luster

0.67

Activations Density 0.035%

words related to negative evaluations or criticisms

contexts that describe inadequately executed actions or conditions

No Comments

No Known Activations