INDEX

Explanations

words related to repeating an action

oai_token-act-pair · gpt-3.5-turbo

the repetition of the word "twice" with varying contexts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Mood

-0.62

FINEST

-0.58

Tags

-0.57

 enactment

-0.57

iation

-0.54

 Discussion

-0.54

Bern

-0.54

Installation

-0.54

aci

-0.53

 stems

-0.53

POSITIVE LOGITS

 twice

3.39

 Twice

2.45

 repeatedly

1.34

 tripled

1.20

 doubled

1.14

 again

1.14

 once

1.08

 quadru

1.07

 double

1.05

once

1.04

Activations Density 0.009%

words related to repeating an action

the repetition of the word "twice" with varying contexts

No Comments

No Known Activations