INDEX

Explanations

numbers and quantities in different contexts

oai_token-act-pair · gpt-3.5-turbo

phrases indicating numerical or statistical trends and changes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 dime

-0.74

 sweat

-0.73

 crap

-0.72

 VIDEOS

-0.71

 recess

-0.66

 shit

-0.66

 messenger

-0.64

 buggy

-0.63

 plur

-0.63

 tire

-0.62

POSITIVE LOGITS

 totals

0.84

OTAL

0.81

arthed

0.80

ivably

0.77

phabet

0.75

oso

0.73

 total

0.72

 Estimated

0.69

Moore

0.68

odiac

0.67

Activations Density 0.241%

numbers and quantities in different contexts

phrases indicating numerical or statistical trends and changes

No Comments

No Known Activations