INDEX

Explanations

references to various items or actions that are related or connected to other specific items or actions

oai_token-act-pair · gpt-3.5-turbo

conjunctions indicating connections and relationships between ideas

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

ito

-0.60

 pian

-0.59

 assassins

-0.58

 Trip

-0.58

 aggro

-0.57

%%

-0.57

 appro

-0.57

obos

-0.56

 gunmen

-0.56

 microphones

-0.56

POSITIVE LOGITS

worldly

0.78

chance

0.67

mentioned

0.64

 Helpful

0.61

cture

0.60

 Pastebin

0.60

enabled

0.59

haven

0.59

fal

0.58

ALE

0.57

Activations Density 0.159%

references to various items or actions that are related or connected to other specific items or actions

conjunctions indicating connections and relationships between ideas

No Comments

No Known Activations