INDEX

Explanations

occurrences of the word "Par" followed by a number

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Par" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

éĹĺ

-0.89

ħĭ

-0.86

 Carbuncle

-0.82

ģĸ

-0.76

hirt

-0.75

OME

-0.75

¥µ

-0.71

ARDS

-0.71

sonian

-0.70

å£«

-0.70

POSITIVE LOGITS

agraph

1.30

anoia

1.17

allel

1.12

vati

1.06

liament

1.01

alysis

1.01

cel

0.96

amount

0.94

abolic

0.93

anthrop

0.92

Activations Density 0.015%

occurrences of the word "Par" followed by a number

occurrences of the word "Par" in various contexts

No Comments

No Known Activations