INDEX

Explanations

proper nouns related to names

oai_token-act-pair · gpt-3.5-turbo

instances of the term "sha" and related variations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 McCann

-0.92

gdala

-0.85

kson

-0.83

entric

-0.81

lio

-0.75

anyl

-0.73

ersen

-0.73

lore

-0.73

nesia

-0.73

ktop

-0.72

POSITIVE LOGITS

SHA

0.96

SHA

0.93

0.88

WARE

0.78

IELD

0.75

Ø¯

0.75

Hash

0.74

Tsu

0.74

âĺħâĺħ

0.72

pless

0.71

Activations Density 0.015%

proper nouns related to names

instances of the term "sha" and related variations

No Comments

No Known Activations