INDEX

Explanations

terms related to coding, such as code snippets, variable names, and programming concepts

oai_token-act-pair · gpt-3.5-turbo

programming terms and syntax errors

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

conservancy

-0.91

terday

-0.88

abouts

-0.76

âĢİ

-0.75

theless

-0.72

ãĤ£

-0.71

zyk

-0.69

football

-0.66

ylum

-0.66

jury

-0.65

POSITIVE LOGITS

Output

1.03

Input

0.94

Node

0.93

String

0.92

Arg

0.90

 Output

0.89

Context

0.88

 Context

0.85

API

0.85

Array

0.84

Activations Density 0.185%

terms related to coding, such as code snippets, variable names, and programming concepts

programming terms and syntax errors

No Comments

No Known Activations