INDEX

Explanations

programming-related terms and syntax sequences

oai_token-act-pair · gpt-3.5-turbo

less-than signs in programming code

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ModLoader

-0.86

ãĥ£

-0.80

 externalToEVAOnly

-0.70

rador

-0.69

Ø¯

-0.69

 Leban

-0.69

 conclud

-0.68

wagen

-0.67

 horizont

-0.66

oise

-0.66

POSITIVE LOGITS

lambda

0.99

span

0.95

_>

0.93

std

0.91

!--

0.87

functional

0.81

sup

0.80

insert

0.79

meta

0.78

chron

0.75

Activations Density 0.009%

programming-related terms and syntax sequences

less-than signs in programming code

No Comments

No Known Activations