INDEX

Explanations

words related to legal processes or legal actions

oai_token-act-pair · gpt-3.5-turbo

mentions of "trauma" or related concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.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

 Mellon

-0.78

 retention

-0.72

hyde

-0.72

ensor

-0.69

 equity

-0.67

 Equity

-0.62

ifference

-0.61

etts

-0.61

ãģ®éŃĶ

-0.61

ENSE

-0.60

POSITIVE LOGITS

pez

0.88

vest

0.87

vert

0.85

ppings

0.85

ph

0.81

fact

0.81

================================================================

0.78

tty

0.78

ordinary

0.78

inen

0.76

Activations Density 0.023%

words related to legal processes or legal actions

mentions of "trauma" or related concepts

No Comments

No Known Activations