INDEX

Explanations

adjectives that express various degrees of intensity, mostly related to negative or critical language

oai_token-act-pair · gpt-3.5-turbo

adjectives that describe varying degrees of quality, often in a critical or evaluative context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ĸļ

-1.10

ICLE

-0.83

atever

-0.72

 TAMADRA

-0.70

heid

-0.69

 Uriel

-0.66

 Happiness

-0.66

 Debor

-0.65

ACTED

-0.63

emale

-0.63

POSITIVE LOGITS

igans

0.83

worldly

0.78

legged

0.73

uitous

0.71

 caveats

0.66

 surprises

0.66

place

0.66

 flair

0.65

 touches

0.65

 shenan

0.65

Activations Density 0.182%

adjectives that express various degrees of intensity, mostly related to negative or critical language

adjectives that describe varying degrees of quality, often in a critical or evaluative context

No Comments

No Known Activations