INDEX

Explanations

words related to streamlining processes or systems

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "line" 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 @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

rum

-0.84

sburg

-0.76

etsy

-0.73

ateurs

-0.72

CVE

-0.72

Reviewer

-0.70

 Seym

-0.70

IRD

-0.69

rito

-0.69

rament

-0.68

POSITIVE LOGITS

backer

1.16

lined

0.94

lines

0.86

lihood

0.78

wise

0.76

breaks

0.75

line

0.72

xual

0.72

break

0.70

xus

0.70

Activations Density 0.015%

words related to streamlining processes or systems

occurrences of the word "line" in various contexts

No Comments

No Known Activations