INDEX

Explanations

proper nouns related to various names and locations

oai_token-act-pair · gpt-3.5-turbo

names of people and locations, particularly focusing on first and last names

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

 claimants

-0.75

adish

-0.72

ulates

-0.66

tty

-0.65

 coma

-0.65

 necessities

-0.65

TY

-0.64

rants

-0.63

 cumbers

-0.62

POSITIVE LOGITS

neau

0.91

spir

0.76

qua

0.74

ski

0.72

claimer

0.72

 Franch

0.71

EEE

0.70

als

0.69

nered

0.68

unct

0.68

Activations Density 0.045%

proper nouns related to various names and locations

names of people and locations, particularly focusing on first and last names

No Comments

No Known Activations