INDEX

Explanations

keywords related to the order or sequence of information

oai_token-act-pair · gpt-3.5-turbo

ordinal numbers indicating a sequence or ranking

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

onomic

-0.66

 budgets

-0.62

itability

-0.59

 budget

-0.57

bind

-0.57

onom

-0.55

 arsen

-0.55

im

-0.55

 irresist

-0.55

activ

-0.54

POSITIVE LOGITS

theless

0.91

 caveat

0.78

named

0.72

foundland

0.71

ngth

0.71

 secondly

0.69

 Centauri

0.69

etheless

0.69

ulhu

0.68

isode

0.65

Activations Density 0.084%

keywords related to the order or sequence of information

ordinal numbers indicating a sequence or ranking

No Comments

No Known Activations