INDEX

Explanations

proper nouns, specifically names of people

oai_token-act-pair · gpt-3.5-turbo

proper nouns and terms related to specific individuals and entertainment content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

llor

-0.85

 Slayer

-0.68

 Dahl

-0.68

 archived

-0.66

ersen

-0.66

UTERS

-0.66

 Hunts

-0.66

owered

-0.65

erers

-0.65

erences

-0.64

POSITIVE LOGITS

forward

0.93

creen

0.89

acies

0.84

nces

0.84

acy

0.84

ahime

0.82

endiary

0.72

Asia

0.71

ourgeois

0.71

law

0.71

Activations Density 0.044%

proper nouns, specifically names of people

proper nouns and terms related to specific individuals and entertainment content

No Comments

No Known Activations