INDEX

Explanations

terms related to religious figures or places in the Middle East

oai_token-act-pair · gpt-3.5-turbo

references to Islamic figures or entities associated with leadership and authority in a political context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 Cosponsors

-0.85

eele

-0.80

ngth

-0.72

ctors

-0.71

hement

-0.69

ilage

-0.68

Reviewer

-0.65

tein

-0.65

ħĭ

-0.65

bably

-0.64

POSITIVE LOGITS

ollah

0.73

hya

0.68

asca

0.68

romy

0.66

NB

0.65

Zen

0.63

Pil

0.61

stan

0.59

mad

0.58

bor

0.58

Activations Density 0.163%

terms related to religious figures or places in the Middle East

references to Islamic figures or entities associated with leadership and authority in a political context

No Comments

No Known Activations