INDEX

Explanations

keywords related to specific events or topics, likely related to news articles or reports

oai_token-act-pair · gpt-3.5-turbo

terms related to governmental and procedural activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 innocence

-0.56

 disadvant

-0.54

ords

-0.54

licted

-0.54

rers

-0.53

ean

-0.52

forcing

-0.52

entimes

-0.51

cers

-0.50

igans

-0.50

POSITIVE LOGITS

 âĢº

0.88

 âĢ¢

0.84

 Â·

0.80

↵

0.79

Â¶

0.78

↵↵

0.77

(%)

0.71

<<

0.70

<|endoftext|>

0.69

 Released

0.69

Activations Density 0.692%

keywords related to specific events or topics, likely related to news articles or reports

terms related to governmental and procedural activities

No Comments

No Known Activations