INDEX

Explanations

specific references to well-known people, places, or events

oai_token-act-pair · gpt-3.5-turbo

references to notable or well-known entities or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

THER

-0.77

atever

-0.77

levant

-0.76

ilation

-0.75

joice

-0.75

xy

-0.73

moving

-0.73

izons

-0.73

Ģ

-0.73

soon

-0.71

POSITIVE LOGITS

 quote

0.91

 landmarks

0.79

 quotation

0.78

 Circus

0.77

 quotes

0.76

aph

0.75

 phrase

0.74

 photograph

0.74

 corrid

0.73

 tale

0.73

Activations Density 0.067%

specific references to well-known people, places, or events

references to notable or well-known entities or concepts

No Comments

No Known Activations