INDEX

Explanations

terms related to statistics or numerical values

oai_token-act-pair · gpt-3.5-turbo

words related to statistics and numerical data

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

lease

-0.77

aunder

-0.73

hops

-0.73

 Karen

-0.71

 Sinai

-0.71

 Barker

-0.70

 Unified

-0.69

zos

-0.69

earchers

-0.68

bara

-0.68

POSITIVE LOGITS

 stat

3.84

stat

2.04

Stat

1.89

 stats

1.84

 Stat

1.84

 statistic

1.61

 STAT

1.37

STAT

1.34

 Stats

1.27

Stats

1.25

Activations Density 0.015%

terms related to statistics or numerical values

words related to statistics and numerical data

No Comments

No Known Activations