INDEX

Explanations

references to specific names, places, and organizations, particularly in the context of news articles

oai_token-act-pair · gpt-3.5-turbo

proper nouns and references to people or organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

.).

-0.74

onto

-0.68

Compat

-0.62

)).

-0.62

Redditor

-0.61

Located

-0.57

'.

-0.56

 unstoppable

-0.56

 respectively

-0.55

atever

-0.53

POSITIVE LOGITS

NET

0.62

enza

0.59

 nutshell

0.58

moon

0.55

endum

0.54

 Explan

0.53

imester

0.53

urious

0.53

oneliness

0.53

 explanations

0.52

Activations Density 2.012%

references to specific names, places, and organizations, particularly in the context of news articles

proper nouns and references to people or organizations

No Comments

No Known Activations