INDEX

Explanations

words related to specific locations or names

oai_token-act-pair · gpt-3.5-turbo

certain names and terms associated with specific places or individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Aircraft

-0.81

 flies

-0.80

 Tail

-0.78

ndra

-0.75

planes

-0.73

Ign

-0.72

 Motorola

-0.70

 Turbo

-0.70

Ign

-0.70

 trainers

-0.70

POSITIVE LOGITS

 Cumber

2.87

anse

1.72

hene

1.53

 Hamp

1.28

Nad

1.21

ansson

1.07

fro

1.06

 Hobbit

1.04

enstein

1.02

Kas

1.02

Activations Density 0.058%

words related to specific locations or names

certain names and terms associated with specific places or individuals

No Comments

No Known Activations