INDEX

Explanations

references or mentions in a text

oai_token-act-pair · gpt-3.5-turbo

references to ongoing situations or conflicts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ummy

-0.72

ternity

-0.70

cmp

-0.68

 bounty

-0.65

pec

-0.65

ont

-0.62

 treasure

-0.62

peat

-0.62

atron

-0.61

oen

-0.61

POSITIVE LOGITS

heads

0.78

 Pitch

0.77

apore

0.76

LY

0.70

sonian

0.66

assis

0.63

 Franch

0.63

=]

0.62

 aside

0.62

Kos

0.61

Activations Density 0.163%

references or mentions in a text

references to ongoing situations or conflicts

No Comments

No Known Activations