INDEX

Explanations

variables and code snippets, including assigning values and functions

oai_token-act-pair · gpt-3.5-turbo

elements related to programming syntax and data structures

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

 oppos

-0.73

pmwiki

-0.71

arily

-0.57

 ACTIONS

-0.57

taboola

-0.56

 adversaries

-0.55

 incumb

-0.54

 conflic

-0.54

allel

-0.54

judicial

-0.53

POSITIVE LOGITS

Shell

0.62

KL

0.57

 Pass

0.55

 McKay

0.54

 Shell

0.54

 Channel

0.54

example

0.53

 Label

0.53

 Example

0.52

Jr

0.51

Activations Density 1.222%

variables and code snippets, including assigning values and functions

elements related to programming syntax and data structures

No Comments

No Known Activations