INDEX

Explanations

words related to computer programming and syntax

oai_token-act-pair · gpt-3.5-turbo

various suffixes and word endings commonly used in English

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

accompan

-0.75

 Instr

-0.66

 Pric

-0.65

 Highlands

-0.65

emale

-0.62

 Seat

-0.62

 Channel

-0.58

è¦ļéĨĴ

-0.58

 Buddy

-0.58

 corrid

-0.58

POSITIVE LOGITS

mosp

0.77

chwitz

0.71

lement

0.71

clair

0.70

verage

0.67

gment

0.67

orm

0.65

phal

0.65

pect

0.64

ircraft

0.64

Activations Density 0.072%

words related to computer programming and syntax

various suffixes and word endings commonly used in English

No Comments

No Known Activations