INDEX

Explanations

various instances of the word "identify"

oai_token-act-pair · gpt-3.5-turbo

phrases related to the act of identification or determining specifics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

orld

-0.86

fare

-0.76

imeo

-0.71

perty

-0.71

bare

-0.65

 raining

-0.64

hands

-0.64

 beating

-0.63

 acquitted

-0.62

ework

-0.62

POSITIVE LOGITS

ãĤ¿

0.79

ively

0.78

ãĤ©

0.74

 Disorders

0.71

 è£ıè

0.70

pointers

0.69

 Ident

0.69

atives

0.68

Ident

0.68

ibly

0.68

Activations Density 0.021%

various instances of the word "identify"

phrases related to the act of identification or determining specifics

No Comments

No Known Activations