INDEX

Explanations

phrases containing vulgar language

oai_token-act-pair · gpt-3.5-turbo

expressions of frustration and disdain

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

earchers

-0.83

Enlarge

-0.77

 envis

-0.77

 vividly

-0.73

 strikingly

-0.70

 outset

-0.69

 foresee

-0.69

 sequencing

-0.67

 locating

-0.66

 doubtless

-0.66

POSITIVE LOGITS

 shit

1.59

 shitty

1.55

 FUCK

1.52

 fucking

1.44

fuck

1.44

Fuck

1.44

shit

1.41

 retarded

1.39

 fuckin

1.38

 fuck

1.35

Activations Density 0.889%

phrases containing vulgar language

expressions of frustration and disdain

No Comments

No Known Activations