INDEX

Explanations

references to physical objects like bands, whether they are related to music, medical usage, or other contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "band."

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

aminer

-0.66

 IMAGES

-0.63

ensional

-0.62

ultz

-0.58

srf

-0.58

bably

-0.57

HY

-0.56

 Toll

-0.55

zanne

-0.55

Hus

-0.55

POSITIVE LOGITS

leader

1.12

width

1.11

camp

1.07

wagon

1.02

ictionary

0.99

band

0.95

ages

0.91

mates

0.89

bands

0.88

ittle

0.87

Activations Density 0.021%

references to physical objects like bands, whether they are related to music, medical usage, or other contexts

instances of the word "band."

No Comments

No Known Activations