INDEX

Explanations

phrases related to adaptations or versions of something from one form to another

oai_token-act-pair · gpt-3.5-turbo

references to adaptations and ports of literature or media

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

 glare

-0.72

hap

-0.67

 disadvantage

-0.66

 refuel

-0.66

Connector

-0.63

 exercised

-0.63

Ku

-0.62

 stripe

-0.61

 shields

-0.61

enged

-0.61

POSITIVE LOGITS

 anthology

0.97

 classics

0.97

 albums

0.92

 remake

0.91

 excerpts

0.91

 installments

0.87

 sequels

0.87

 compilation

0.86

 autobiography

0.85

 adaptation

0.85

Activations Density 0.662%

phrases related to adaptations or versions of something from one form to another

references to adaptations and ports of literature or media

No Comments

No Known Activations