INDEX

Explanations

instances of the word "no" used within various contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "no" and phrases expressing refusal or denial

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

otle

-0.68

lycer

-0.68

ivic

-0.66

ortment

-0.66

ulative

-0.64

 rall

-0.64

tyard

-0.64

 Halls

-0.62

nox

-0.62

worn

-0.62

POSITIVE LOGITS

zzle

0.89

etheless

0.84

ICE

0.83

terday

0.79

xious

0.79

isy

0.77

AH

0.74

 kidding

0.71

esis

0.70

ct

0.69

Activations Density 0.063%

instances of the word "no" used within various contexts

instances of the word "no" and phrases expressing refusal or denial

No Comments

No Known Activations