INDEX

Explanations

phrases related to performing actions repeatedly or in sequence

oai_token-act-pair · gpt-3.5-turbo

phrases related to sequential actions or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

erville

-0.68

THING

-0.66

Sport

-0.65

ura

-0.62

WI

-0.61

ubb

-0.59

ª

-0.59

ļ

-0.58

ergy

-0.57

 Shame

-0.56

POSITIVE LOGITS

 consecut

1.15

 depending

1.05

 spaced

1.03

 horizontally

1.00

 across

0.96

 eliminating

0.93

 concurrently

0.90

 throughout

0.90

 simultaneously

0.89

 alternating

0.88

Activations Density 0.327%

phrases related to performing actions repeatedly or in sequence

phrases related to sequential actions or processes

No Comments

No Known Activations