INDEX

Explanations

any mentions of specific grades

oai_token-act-pair · gpt-3.5-turbo

references to academic grading systems

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

 Bundy

-0.90

ĸļ

-0.74

eers

-0.73

orial

-0.72

iak

-0.71

reprene

-0.69

thia

-0.69

ogene

-0.68

riks

-0.68

issance

-0.68

POSITIVE LOGITS

 grades

0.99

 rating

0.89

 grade

0.89

 grading

0.85

Grade

0.78

 plutonium

0.76

ably

0.75

 ratings

0.72

 insign

0.68

ific

0.68

Activations Density 0.016%

any mentions of specific grades

references to academic grading systems

No Comments

No Known Activations