INDEX

Explanations

Double-letter sequences

oai_token-act-pair · gpt-3.5-turbo

instances of the syllable "bu" or its variations in words

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

alf

-0.79

odor

-0.71

lake

-0.69

oly

-0.69

ashtra

-0.68

opath

-0.68

 Odyssey

-0.67

rums

-0.67

stadt

-0.62

ribution

-0.62

POSITIVE LOGITS

etooth

0.89

gged

0.87

glers

0.86

enos

0.83

gging

0.81

querque

0.81

 dormant

0.79

pt

0.75

tip

0.75

holes

0.75

Activations Density 0.027%

Double-letter sequences

instances of the syllable "bu" or its variations in words

No Comments

No Known Activations