INDEX

Explanations

phrases related to consecutive occurrences or sequences

oai_token-act-pair · gpt-3.5-turbo

references to consecutive or continuous sequences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

adle

-0.80

 Lauder

-0.72

orage

-0.69

mble

-0.65

healthy

-0.62

-0.61

mur

-0.60

akings

-0.60

alg

-0.59

Bog

-0.58

POSITIVE LOGITS

 straight

0.84

ened

0.79

away

0.78

 dope

0.77

VERT

0.75

lined

0.74

bent

0.74

 Straight

0.73

forward

0.73

straight

0.73

Activations Density 0.007%

phrases related to consecutive occurrences or sequences

references to consecutive or continuous sequences

No Comments

No Known Activations