INDEX

Explanations

the number "two" occurring in a sequence

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "two."

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

 Changed

-0.71

rawdownloadcloneembedreportprint

-0.67

trap

-0.66

vale

-0.66

VIDIA

-0.62

bos

-0.62

 discriminating

-0.61

nect

-0.59

 Limits

-0.59

racuse

-0.58

POSITIVE LOGITS

 paragraphs

0.84

teen

0.79

teenth

0.79

aciously

0.78

 installments

0.77

 months

0.74

 digits

0.73

 weeks

0.72

 thirds

0.71

isd

0.71

Activations Density 0.057%

the number "two" occurring in a sequence

occurrences of the word "two."

No Comments

No Known Activations