INDEX

Explanations

years

oai_token-act-pair · gpt-3.5-turbo

specific years mentioned in a historical context

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

amy

-0.88

thia

-0.69

umers

-0.68

adata

-0.65

 ashore

-0.64

merce

-0.63

terday

-0.63

ksh

-0.61

olen

-0.61

washing

-0.61

POSITIVE LOGITS

 edition

1.00

 iteration

0.80

0.79

 version

0.77

 Confeder

0.75

 installment

0.72

 incident

0.71

 timeframe

0.70

 annual

0.70

 Annual

0.70

Activations Density 0.110%

years

specific years mentioned in a historical context

No Comments

No Known Activations