INDEX

Explanations

specific names or identifiers within code snippets

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "name" and its variations, suggesting a focus on identifiers or naming conventions in a programming context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

romy

-0.86

yrinth

-0.80

ilib

-0.77

icult

-0.72

 balcon

-0.70

awaru

-0.69

avorite

-0.68

DERR

-0.66

psey

-0.65

izoph

-0.65

POSITIVE LOGITS

plates

1.14

paces

1.04

name

1.03

plate

1.00

names

1.00

ames

0.90

Name

0.84

 NAME

0.82

="

0.81

hood

0.80

Activations Density 0.027%

specific names or identifiers within code snippets

occurrences of the word "name" and its variations, suggesting a focus on identifiers or naming conventions in a programming context

No Comments

No Known Activations