INDEX

Explanations

references to the term "Moore"

oai_token-act-pair · gpt-3.5-turbo

the recurring mention of the name "Moore."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ERAL

-0.84

ropolitan

-0.82

ICAN

-0.73

ivity

-0.73

æĺ¯

-0.68

Spanish

-0.66

===

-0.66

rious

-0.66

 Hungarian

-0.66

lift

-0.65

POSITIVE LOGITS

 Moore

1.15

Moore

0.91

stown

0.79

oshenko

0.75

ufact

0.75

cale

0.75

acre

0.74

zilla

0.74

sey

0.73

cloth

0.73

Activations Density 0.006%

references to the term "Moore"

the recurring mention of the name "Moore."

No Comments

No Known Activations