INDEX

Explanations

corrections or editorial notes within text

oai_token-act-pair · gpt-3.5-turbo

instances of the word "correction" and its variations

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

atos

-0.94

bang

-0.76

axy

-0.72

ramid

-0.69

ulhu

-0.68

gha

-0.67

aden

-0.65

WAYS

-0.65

ahan

-0.64

enf

-0.64

POSITIVE LOGITS

 inaccur

0.77

 leveled

0.76

Correction

0.74

Reviewed

0.71

 Corrections

0.69

ettings

0.68

 Correction

0.66

 empt

0.66

Correct

0.66

 inaccurate

0.65

Activations Density 0.027%

corrections or editorial notes within text

instances of the word "correction" and its variations

No Comments

No Known Activations