INDEX

Explanations

coding-related terms and technical concepts

oai_token-act-pair · gpt-3.5-turbo

technical terms related to programming and system functionalities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

bane

-0.94

mare

-0.87

-0.86

bent

-0.81

bill

-0.80

mast

-0.80

psy

-0.80

deck

-0.78

ende

-0.78

mort

-0.77

POSITIVE LOGITS

 sandbox

0.89

 defaults

0.85

 workaround

0.85

 login

0.81

 runtime

0.80

 deprecated

0.80

API

0.79

 namespace

0.79

 parsing

0.78

 syntax

0.78

Activations Density 0.579%

coding-related terms and technical concepts

technical terms related to programming and system functionalities

No Comments

No Known Activations