INDEX

Explanations

time-related events that happened in the past

oai_token-act-pair · gpt-3.5-turbo

references to time, specifically events that occurred in the past

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

 Scully

-0.66

nce

-0.64

 Cipher

-0.64

 Scal

-0.64

Sph

-0.61

Gad

-0.59

icular

-0.59

 Stellar

-0.58

 Trou

-0.57

anium

-0.57

POSITIVE LOGITS

 earthqu

0.85

 mosqu

0.82

 apiece

0.81

 nomine

0.76

 nodd

0.73

apon

0.71

nown

0.70

 eleph

0.70

 conflic

0.68

ivari

0.65

Activations Density 0.057%

time-related events that happened in the past

references to time, specifically events that occurred in the past

No Comments

No Known Activations