INDEX

Explanations

specific political references and notable events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.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

 numerous

-0.72

 warranties

-0.71

 Cavern

-0.70

 Ninth

-0.67

AFB

-0.67

 partner

-0.66

 Chao

-0.66

 Goodwin

-0.64

 Lair

-0.64

 tantal

-0.64

POSITIVE LOGITS

1.05

ghazi

0.96

agate

0.95

olitics

0.94

Leaks

0.91

__

0.89

#$#$

0.89

India

0.89

wcs

0.86

Activations Density 0.027%

specific political references and notable events

No Comments

No Known Activations