INDEX

Explanations

occurrences of the word "and"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

lite

-0.95

icut

-0.79

 ACLU

-0.69

lio

-0.66

payer

-0.65

aign

-0.64

cum

-0.64

urses

-0.63

aun

-0.62

appointed

-0.61

POSITIVE LOGITS

 unanswered

0.66

 attitude

0.65

 Hayward

0.63

 teamwork

0.62

 hemisphere

0.61

berra

0.61

ãĤ£

0.60

 diseng

0.59

acity

0.59

hered

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.