INDEX

Explanations

words or parts of words related to physical actions or processes

oai_token-act-pair · gpt-3.5-turbo

past and present participles of verbs

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

kefeller

-0.79

displayText

-0.70

 actionGroup

-0.70

 Leilan

-0.70

bnb

-0.70

ITNESS

-0.69

unic

-0.63

ãĤ©

-0.63

TION

-0.63

è£ħ

-0.62

POSITIVE LOGITS

 breaths

0.69

 herself

0.66

 fide

0.65

ĺħ

0.64

kees

0.60

tons

0.60

atio

0.60

ings

0.60

 lett

0.59

borough

0.58

Activations Density 0.084%

words or parts of words related to physical actions or processes

past and present participles of verbs

No Comments

No Known Activations