INDEX

Explanations

words related to names or specific terms, particularly those with the pattern "P...l"

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names and locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

schild

-0.84

ATTLE

-0.63

 quantity

-0.61

 Frames

-0.57

 Whole

-0.56

 Wonderland

-0.56

 Clover

-0.56

 Frequ

-0.56

 Raven

-0.55

 striking

-0.55

POSITIVE LOGITS

ongyang

1.13

sylvania

0.87

asant

0.81

asus

0.79

formance

0.77

haps

0.77

chedel

0.76

eworthy

0.76

sburgh

0.75

ucket

0.74

Activations Density 0.091%

words related to names or specific terms, particularly those with the pattern "P...l"

proper nouns, particularly names and locations

No Comments

No Known Activations