INDEX

Explanations

names of individuals or entities in different contexts

oai_token-act-pair · gpt-3.5-turbo

verbs associated with actions or activities

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

ogether

-0.60

 equivalents

-0.58

 equivalent

-0.55

sylv

-0.54

selves

-0.54

 corrid

-0.53

results

-0.52

Compar

-0.51

halla

-0.51

 referen

-0.50

POSITIVE LOGITS

bra

0.58

 himself

0.54

hirt

0.53

itone

0.53

 Expand

0.50

his

0.49

NHL

0.49

TOR

0.48

brew

0.48

UFC

0.47

Activations Density 0.839%

names of individuals or entities in different contexts

verbs associated with actions or activities

No Comments

No Known Activations