INDEX

Explanations

names or terms related to computer programming languages and file extensions

oai_token-act-pair · gpt-3.5-turbo

variables or terms used in programming code and file naming conventions

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

 Governors

-0.74

 Samar

-0.72

 Beir

-0.71

 Democracy

-0.68

Eag

-0.63

 governors

-0.63

 sweeps

-0.62

 filibuster

-0.62

TBA

-0.61

 electing

-0.61

POSITIVE LOGITS

cdn

1.16

rc

0.95

tml

0.95

ignore

0.91

jc

0.87

/.

0.86

aml

0.86

#$

0.86

xml

0.86

igslist

0.84

Activations Density 0.082%

names or terms related to computer programming languages and file extensions

variables or terms used in programming code and file naming conventions

No Comments

No Known Activations