INDEX

Explanations

names or words related to specific individuals

oai_token-act-pair · gpt-3.5-turbo

instances of the word "ind" in various contexts or forms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 tremend

-0.73

 compr

-0.69

©¶æ

-0.69

Ĥİ

-0.69

senal

-0.68

 litter

-0.68

EStream

-0.66

ascus

-0.65

awaru

-0.64

 rubbish

-0.63

POSITIVE LOGITS

irect

1.15

icator

1.05

icative

0.97

ividually

0.96

sight

0.94

ications

0.92

icators

0.89

ebted

0.88

icates

0.87

ication

0.87

Activations Density 0.009%

names or words related to specific individuals

instances of the word "ind" in various contexts or forms

No Comments

No Known Activations