INDEX

Explanations

terms related to locations or specific positions within a setting

oai_token-act-pair · gpt-3.5-turbo

specific references to objects, locations, and technological components

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

Pitt

-0.74

 abst

-0.65

DragonMagazine

-0.64

AAA

-0.62

 improv

-0.61

Ô

-0.61

Ïģ

-0.61

rosso

-0.61

cific

-0.59

————

-0.59

POSITIVE LOGITS

 drawer

1.03

 cabinets

0.98

 bloodstream

0.94

 throats

0.93

 cavity

0.90

 cylinders

0.89

 ceiling

0.89

 shelves

0.88

 pedest

0.86

 doorway

0.85

Activations Density 0.821%

terms related to locations or specific positions within a setting

specific references to objects, locations, and technological components

No Comments

No Known Activations