INDEX

Explanations

phrases related to different ways or methods

oai_token-act-pair · gpt-3.5-turbo

phrases that describe various methods or ways of doing something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

bryce

-0.64

fleet

-0.61

ŃĶ

-0.57

roma

-0.56

 SOLD

-0.56

OPA

-0.55

RA

-0.54

 Lear

-0.52

 worried

-0.52

IOR

-0.51

POSITIVE LOGITS

 which

1.23

 Which

0.96

ahime

0.93

which

0.93

mental

0.76

 whereby

0.74

Which

0.74

effic

0.70

 whence

0.69

 guiActiveUnfocused

0.68

Activations Density 0.061%

phrases related to different ways or methods

phrases that describe various methods or ways of doing something

No Comments

No Known Activations