INDEX

Explanations

mentions of the abbreviation "CM."

oai_token-act-pair · gpt-3.5-turbo

references to a political figure or leadership role identified by the acronym "CM"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

glers

-0.87

furt

-0.79

gers

-0.78

 sidx

-0.76

via

-0.75

ibility

-0.75

swick

-0.74

âĸĪâĸĪ

-0.73

heid

-0.71

get

-0.70

POSITIVE LOGITS

CM

1.00

ODE

0.89

CM

0.86

ipop

0.84

CP

0.82

ategory

0.81

ancel

0.77

oly

0.74

oker

0.73

OS

0.72

Activations Density 0.003%

mentions of the abbreviation "CM."

references to a political figure or leadership role identified by the acronym "CM"

No Comments

No Known Activations