INDEX

Explanations

proper nouns representing various titles or characters, especially from media and entertainment

oai_token-act-pair · gpt-3.5-turbo

titles or names of various media, particularly games and characters

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

 acknow

-0.95

 confir

-0.82

etheless

-0.79

depending

-0.77

 notor

-0.75

elong

-0.73

perature

-0.72

 recourse

-0.71

 referen

-0.71

 skilled

-0.71

POSITIVE LOGITS

 Productions

1.15

 Studios

1.02

 Syndrome

0.97

 Remix

0.95

 Massacre

0.94

 Collection

0.93

 Blues

0.92

 Circus

0.91

 Brigade

0.91

 Squad

0.90

Activations Density 0.301%

proper nouns representing various titles or characters, especially from media and entertainment

titles or names of various media, particularly games and characters

No Comments

No Known Activations