INDEX

Explanations

patterns related to programming code

oai_token-act-pair · gpt-3.5-turbo

programming syntax and variable definitions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

SPONSORED

-1.32

impro

-1.07

Advertisements

-1.07

Poor

-1.07

ESA

-1.06

Nap

-1.04

AZ

-1.04

Extreme

-1.04

"âĢ¦

-1.03

Prince

-1.03

POSITIVE LOGITS

ptr

1.08

buf

1.02

 addr

1.02

ml

1.00

cf

1.00

 append

0.97

seq

0.97

srf

0.97

 lang

0.97

 cursor

0.96

Activations Density 0.076%

patterns related to programming code

programming syntax and variable definitions

No Comments

No Known Activations