INDEX

Explanations

references to movies

oai_token-act-pair · gpt-3.5-turbo

references to movies

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 distress

-0.73

 stressing

-0.69

 pains

-0.69

orns

-0.67

rist

-0.66

ashington

-0.66

nesota

-0.65

 churn

-0.65

 commission

-0.64

 dread

-0.64

POSITIVE LOGITS

 Movie

3.81

Movie

2.68

 Movies

2.46

movie

1.96

 Cinema

1.80

 Film

1.69

 Films

1.51

 Cinem

1.51

 Animation

1.50

Film

1.48

Activations Density 0.026%

references to movies

references to movies

No Comments

No Known Activations