INDEX

Explanations

personal names and surnames, potentially of public figures

oai_token-act-pair · gpt-3.5-turbo

names of people or personalities

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

intendent

-0.75

Reviewer

-0.73

stown

-0.72

ruary

-0.72

Äĩ

-0.72

rawdownloadcloneembedreportprint

-0.70

dylib

-0.70

 Hurricanes

-0.67

LIA

-0.65

inations

-0.65

POSITIVE LOGITS

isner

0.67

ravis

0.62

amoto

0.60

asma

0.60

 vacuum

0.59

acher

0.57

 lawy

0.57

ioxide

0.57

İ

0.56

enty

0.56

Activations Density 0.228%

personal names and surnames, potentially of public figures

names of people or personalities

No Comments

No Known Activations