INDEX

Explanations

phrases indicating that something is under consideration or in progress

oai_token-act-pair · gpt-3.5-turbo

phrases related to items that are being discussed or considered

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

raped

-0.73

IZE

-0.71

Ì¶

-0.69

ISON

-0.69

irection

-0.68

RANT

-0.68

OME

-0.67

 Invalid

-0.67

itism

-0.63

arson

-0.63

POSITIVE LOGITS

 horizon

1.38

 agenda

1.18

 chopping

1.12

 roadmap

1.11

 shelf

1.10

 shelves

1.08

 table

1.08

 doorstep

1.05

 minds

1.02

 menu

0.98

Activations Density 0.091%

phrases indicating that something is under consideration or in progress

phrases related to items that are being discussed or considered

No Comments

No Known Activations