INDEX

Explanations

phrases related to titles or subjects of artistic works

oai_token-act-pair · gpt-3.5-turbo

song titles and lyrics that convey themes of hope, love, and resilience

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 sidew

-0.69

 selective

-0.68

 refining

-0.66

 deleg

-0.66

 popul

-0.65

 scaven

-0.65

auder

-0.65

 forestry

-0.65

anx

-0.64

 strengths

-0.64

POSITIVE LOGITS

 Wrong

1.11

 Hate

1.05

 Feeling

1.02

 Happ

1.00

 Tonight

0.98

 Changed

0.98

 Called

0.98

 Bastard

0.97

 Forever

0.97

 Goodbye

0.97

Activations Density 0.263%

phrases related to titles or subjects of artistic works

song titles and lyrics that convey themes of hope, love, and resilience

No Comments

No Known Activations