INDEX

Explanations

words related to laps or lapses in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to lapses in attention or memory

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

 Square

-0.69

 Antar

-0.64

 Towns

-0.63

Square

-0.62

 Âł Âł Âł Âł Âł Âł Âł Âł

-0.61

GROUP

-0.61

 prisoner

-0.60

BI

-0.59

ISON

-0.59

 Âł Âł Âł Âł

-0.58

POSITIVE LOGITS

es

1.05

apon

0.86

imet

0.86

lap

0.83

atile

0.83

 laps

0.82

esian

0.81

pit

0.81

isec

0.81

lyak

0.80

Activations Density 0.013%

words related to laps or lapses in various contexts

references to lapses in attention or memory

No Comments

No Known Activations