INDEX

Explanations

instances of events or conditions that happened

oai_token-act-pair · gpt-3.5-turbo

references to various types of cases or incidents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

sonian

-0.77

aye

-0.72

asus

-0.70

externalActionCode

-0.65

isse

-0.65

issance

-0.64

phia

-0.63

lasses

-0.62

vation

-0.61

bern

-0.60

POSITIVE LOGITS

 involving

1.23

 where

0.92

 cases

0.88

 wherein

0.87

hots

0.87

 whereby

0.84

 documented

0.83

 arising

0.82

of

0.81

 occurring

0.81

Activations Density 0.097%

instances of events or conditions that happened

references to various types of cases or incidents

No Comments

No Known Activations