INDEX

Explanations

instances where speculation is being made

oai_token-act-pair · gpt-3.5-turbo

instances of the word "speculate" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

vin

-0.79

mar

-0.77

nea

-0.70

ne

-0.70

nce

-0.69

miss

-0.68

uv

-0.65

anti

-0.65

tha

-0.65

breaking

-0.64

POSITIVE LOGITS

 speculate

1.05

acebook

0.90

 conject

0.86

 speculated

0.83

 theor

0.81

 conjecture

0.80

 FreeBSD

0.77

 guesses

0.74

 spec

0.72

 speculation

0.72

Activations Density 0.016%

instances where speculation is being made

instances of the word "speculate" and its variations

No Comments

No Known Activations