INDEX

Explanations

terms related to software development, particularly around tracking and managing errors and program states

oai_token-act-pair · gpt-3.5-turbo

references to current and previous states or values in a system

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

gging

-0.79

leaders

-0.76

 Schwe

-0.72

Sov

-0.71

NEWS

-0.71

spirit

-0.70

NEY

-0.70

 whistleblowers

-0.69

perse

-0.68

HCR

-0.68

POSITIVE LOGITS

 decimal

0.94

 node

0.94

 ancestor

0.93

(<

0.87

 layer

0.84

 iteration

0.82

 vertex

0.80

 dataset

0.79

 integer

0.79

 occurrence

0.79

Activations Density 0.219%

terms related to software development, particularly around tracking and managing errors and program states

references to current and previous states or values in a system

No Comments

No Known Activations