INDEX

Explanations

words related to organized activities or events

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the word "session" with varying contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

umbai

-0.81

Vote

-0.67

Lev

-0.65

Fight

-0.57

hood

-0.56

BAT

-0.55

EW

-0.54

 underdog

-0.54

 suspic

-0.54

":"/

-0.54

POSITIVE LOGITS

ional

0.94

 session

0.89

 sessions

0.89

ansion

0.83

essions

0.82

isters

0.80

alion

0.78

lain

0.77

abus

0.77

 Sessions

0.76

Activations Density 0.044%

words related to organized activities or events

repeated mentions of the word "session" with varying contexts

No Comments

No Known Activations