INDEX

Explanations

statistics or numerical expressions

oai_token-act-pair · gpt-3.5-turbo

phrases that contain ratios or proportions expressed as "out of" statements

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

cation

-0.67

selves

-0.63

irez

-0.62

ettel

-0.59

lly

-0.58

 Unch

-0.58

ourgeois

-0.58

htaking

-0.57

introdu

-0.57

ymm

-0.56

POSITIVE LOGITS

ten

0.88

0.83

 thirteen

0.82

0.81

0.80

 tens

0.80

0.79

 nine

0.79

 dozens

0.78

0.77

Activations Density 0.057%

statistics or numerical expressions

phrases that contain ratios or proportions expressed as "out of" statements

No Comments

No Known Activations