INDEX

Explanations

instances of the word "probable" or related terms like "probability", "statistical probability", and "probability distributions."

oai_token-act-pair · gpt-3.5-turbo

terms related to probability and evidence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

uart

-0.95

artney

-0.87

helle

-0.86

agra

-0.83

ravings

-0.80

afort

-0.78

ophon

-0.76

lain

-0.73

raved

-0.72

turned

-0.71

POSITIVE LOGITS

hood

0.88

ties

0.85

 occurrence

0.80

 probability

0.80

 scenario

0.77

 guesses

0.76

 assumption

0.76

 inference

0.75

 Cause

0.75

 cause

0.74

Activations Density 0.083%

instances of the word "probable" or related terms like "probability", "statistical probability", and "probability distributions."

terms related to probability and evidence

No Comments

No Known Activations