INDEX

Explanations

phrases indicating encouragement or agreement

oai_token-act-pair · gpt-3.5-turbo

the phrase "get on" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ikan

-0.66

ascript

-0.65

\\\\\\\\

-0.63

ãĤ©

-0.62

ython

-0.61

 Accessed

-0.60

ãĤ§

-0.59

rors

-0.59

inventoryQuantity

-0.58

cpp

-0.57

POSITIVE LOGITS

erous

1.09

shore

1.08

 board

1.06

 behalf

1.00

stage

0.90

sets

0.87

screen

0.87

yx

0.85

top

0.82

etime

0.81

Activations Density 0.092%

phrases indicating encouragement or agreement

the phrase "get on" in various contexts

No Comments

No Known Activations