INDEX

Explanations

specific academic terms or names, potentially related to scholarly activity or research

oai_token-act-pair · gpt-3.5-turbo

references to the term "Sch" or related names

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Grateful

-0.74

 constrained

-0.71

 monarch

-0.69

 censored

-0.67

 subjected

-0.62

 deprived

-0.61

 measured

-0.61

ocally

-0.61

DEN

-0.60

 Korra

-0.60

POSITIVE LOGITS

izoph

1.10

thal

1.07

acher

1.05

midt

1.04

acht

1.03

ulz

0.99

uler

0.97

olars

0.95

atz

0.95

sch

0.94

Activations Density 0.007%

specific academic terms or names, potentially related to scholarly activity or research

references to the term "Sch" or related names

No Comments

No Known Activations