INDEX

Explanations

statements related to facts or factual information

oai_token-act-pair · gpt-3.5-turbo

mentions of "facts" and their importance in various contexts

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

yss

-0.90

isoft

-0.86

osi

-0.83

antha

-0.83

ovo

-0.81

zik

-0.80

leased

-0.80

hod

-0.79

rint

-0.75

interstitial

-0.75

POSITIVE LOGITS

heet

0.96

 facts

0.88

facts

0.87

fulness

0.87

telling

0.86

 inacc

0.79

 pertaining

0.78

 relating

0.72

 surrounding

0.72

fact

0.71

Activations Density 0.016%

statements related to facts or factual information

mentions of "facts" and their importance in various contexts

No Comments

No Known Activations