INDEX

Explanations

commands or executables stored in directories marked with "bin"

oai_token-act-pair · gpt-3.5-turbo

instances of the term "bin" associated with file paths or directories

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

anwhile

-0.75

 Jagu

-0.71

ammy

-0.71

 entrants

-0.64

pter

-0.63

 advant

-0.62

theless

-0.61

plays

-0.60

gypt

-0.60

IRE

-0.60

POSITIVE LOGITS

ocular

1.57

omial

1.41

 Laden

1.19

ned

1.08

ning

0.98

nington

0.96

thood

0.94

stocks

0.93

jamin

0.93

utils

0.86

Activations Density 0.019%

commands or executables stored in directories marked with "bin"

instances of the term "bin" associated with file paths or directories

No Comments

No Known Activations