INDEX

Explanations

proper nouns and text related to locations or events

oai_token-act-pair · gpt-3.5-turbo

prominent conjunctions and significant phrases that contribute to the structure of arguments or statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

assetsadobe

-0.83

 behav

-0.69

SAN

-0.68

 Convers

-0.65

 SERVICE

-0.63

 acknow

-0.63

 Pens

-0.63

Gle

-0.62

 Wast

-0.62

 ende

-0.61

POSITIVE LOGITS

utenberg

0.83

chin

0.82

chid

0.82

sever

0.79

cox

0.77

eq

0.76

inline

0.76

til

0.76

lder

0.75

gger

0.75

Activations Density 0.202%

proper nouns and text related to locations or events

prominent conjunctions and significant phrases that contribute to the structure of arguments or statements

No Comments

No Known Activations