INDEX

Explanations

mentions of stories or narratives

oai_token-act-pair · gpt-3.5-turbo

references to various narratives or stories

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

ividual

-0.84

erate

-0.82

oved

-0.82

foreseen

-0.81

ierrez

-0.80

uters

-0.76

rador

-0.76

activated

-0.76

raviolet

-0.75

essors

-0.75

POSITIVE LOGITS

 tale

1.45

 tales

1.40

 Tales

1.04

 Tale

1.02

tale

0.95

 saga

0.83

 Ragnarok

0.83

 Leviathan

0.82

tell

0.78

 Mania

0.77

Activations Density 0.015%

mentions of stories or narratives

references to various narratives or stories

No Comments

No Known Activations