INDEX

Explanations

keywords related to questioning or doubt

oai_token-act-pair · gpt-3.5-turbo

instances of the word "questioned" or its 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

oiler

-0.87

emetery

-0.73

ergic

-0.70

rites

-0.69

ccording

-0.69

uscript

-0.67

ategory

-0.67

\\\\

-0.66

ëĭ

-0.66

pite

-0.64

POSITIVE LOGITS

naires

1.15

 questioning

0.96

naire

0.93

 questioned

0.89

 questions

0.87

 Questions

0.81

 unanswered

0.79

ively

0.78

 whether

0.77

 probing

0.76

Activations Density 0.026%

keywords related to questioning or doubt

instances of the word "questioned" or its variations

No Comments

No Known Activations