INDEX

Explanations

phrases expressing strong positive emotions, particularly about different things like recipes, songs, and experiences

oai_token-act-pair · gpt-3.5-turbo

expressions of strong personal preferences or affections

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

uthor

-0.86

river

-0.74

ufact

-0.72

 transpired

-0.71

ailable

-0.68

Corp

-0.67

imester

-0.66

 disadvantage

-0.64

ascript

-0.63

Args

-0.62

POSITIVE LOGITS

 dearly

0.86

 uncond

0.76

 idea

0.72

 spont

0.69

 guts

0.67

 reputation

0.63

 hobby

0.63

 outdoors

0.62

 vibe

0.62

 decoration

0.62

Activations Density 0.291%

phrases expressing strong positive emotions, particularly about different things like recipes, songs, and experiences

expressions of strong personal preferences or affections

No Comments

No Known Activations