INDEX

Explanations

words with the root "extrude" or "intrinsic"

oai_token-act-pair · gpt-3.5-turbo

terms related to extraction and intrinsic properties

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

 Columbia

-0.92

peak

-0.84

 Butterfly

-0.80

 Columb

-0.78

 Blitz

-0.77

leck

-0.77

creen

-0.75

 Cups

-0.74

ELL

-0.72

ership

-0.71

POSITIVE LOGITS

 extr

1.33

aneous

1.00

MpServer

0.89

uder

0.88

 unnecess

0.85

hetical

0.81

å¦

0.80

 srfAttach

0.80

uders

0.78

untled

0.76

Activations Density 0.008%

words with the root "extrude" or "intrinsic"

terms related to extraction and intrinsic properties

No Comments

No Known Activations