INDEX

Explanations

words related to absorption and uptake

oai_token-act-pair · gpt-3.5-turbo

terms related to absorption processes

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

FTWARE

-0.65

Correct

-0.62

 Benghazi

-0.60

 exoner

-0.59

 Investigator

-0.59

Egypt

-0.58

 Else

-0.58

 Violence

-0.58

giving

-0.57

 Libya

-0.57

POSITIVE LOGITS

 absorbs

0.91

 absorbed

0.85

iencies

0.84

ptive

0.83

ency

0.83

 absor

0.82

ibles

0.81

encies

0.80

lesi

0.80

 absorbing

0.79

Activations Density 0.013%

words related to absorption and uptake

terms related to absorption processes

No Known Activations