INDEX

Explanations

programming-related terms, particularly focusing on characters and string manipulations

oai_token-act-pair · gpt-3.5-turbo

elements related to programming and syntax

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

HCR

-0.84

ãĢı

-0.77

ndra

-0.75

 detrim

-0.73

asio

-0.73

Monitor

-0.72

Rah

-0.71

hement

-0.70

DEM

-0.69

 Valkyrie

-0.68

POSITIVE LOGITS

 formatting

1.08

 text

1.06

 comma

1.06

 delim

1.04

 syntax

0.99

 alphabet

0.97

 strings

0.97

 parentheses

0.90

acters

0.89

 parsing

0.89

Activations Density 0.223%

programming-related terms, particularly focusing on characters and string manipulations

elements related to programming and syntax

No Comments

No Known Activations