INDEX

Explanations

phrases related to places or entities with the term "Mon" in them

oai_token-act-pair · gpt-3.5-turbo

references to a specific type of creature, likely "monsters" or similar entities

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

ById

-0.70

ngth

-0.70

 linemen

-0.63

 whistleblowers

-0.63

 patents

-0.62

 redundancy

-0.61

Accessory

-0.61

 glim

-0.60

stood

-0.60

Ö¼

-0.60

POSITIVE LOGITS

afort

0.80

liction

0.79

achine

0.74

asing

0.74

obe

0.73

olithic

0.70

theless

0.70

jong

0.70

 Settlement

0.70

agne

0.69

Activations Density 0.086%

phrases related to places or entities with the term "Mon" in them

references to a specific type of creature, likely "monsters" or similar entities

No Comments

No Known Activations