INDEX

Explanations

statistical correlations or likelihoods related to different scenarios or conditions

oai_token-act-pair · gpt-3.5-turbo

phrases discussing the likelihood or probability of certain conditions or outcomes

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

inth

-0.94

ilts

-0.83

zeb

-0.81

ornia

-0.78

illation

-0.77

andan

-0.77

aredevil

-0.76

gian

-0.74

aan

-0.74

 Dash

-0.72

POSITIVE LOGITS

 infer

0.83

 culprit

0.79

 likely

0.78

Magikarp

0.78

 assum

0.75

 millenn

0.75

 underest

0.74

 underestimate

0.74

 guesses

0.72

 surv

0.71

Activations Density 0.022%

statistical correlations or likelihoods related to different scenarios or conditions

phrases discussing the likelihood or probability of certain conditions or outcomes

No Comments

No Known Activations