INDEX

Explanations

instances of words that are followed by punctuation marks

oai_token-act-pair · gpt-3.5-turbo

occurrences of punctuation and words indicating relationships or dependencies

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

frog

-0.69

HEAD

-0.69

Sham

-0.67

JV

-0.67

CRE

-0.66

 Responsibility

-0.60

riors

-0.59

SAY

-0.58

front

-0.58

 Lean

-0.58

POSITIVE LOGITS

uated

1.83

uating

1.54

uates

1.43

uation

1.35

uate

1.30

uations

1.22

uality

1.12

iously

1.10

aneous

1.10

uitous

1.02

Activations Density 0.046%

instances of words that are followed by punctuation marks

occurrences of punctuation and words indicating relationships or dependencies

No Comments

No Known Activations