INDEX

Explanations

words related to small or interesting pieces of information or details

oai_token-act-pair · gpt-3.5-turbo

references to discrete pieces of information or facts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

HAEL

-0.69

PRES

-0.66

sein

-0.64

heid

-0.63

Rush

-0.63

HS

-0.62

anwhile

-0.58

ean

-0.57

respond

-0.57

ADE

-0.57

POSITIVE LOGITS

tid

1.35

bits

0.89

iquette

0.89

\\\\\\\\

0.88

*=-

0.82

âĶĢâĶĢâĶĢâĶĢ

0.82

Tid

0.81

SpaceEngineers

0.79

iest

0.79

shed

0.78

Activations Density 0.006%

words related to small or interesting pieces of information or details

references to discrete pieces of information or facts

No Comments

No Known Activations