INDEX

Explanations

adjectives and nouns relating to logical reasoning

oai_token-act-pair · gpt-3.5-turbo

logical reasoning and arguments

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

lain

-0.86

adal

-0.75

emi

-0.72

chuk

-0.68

rael

-0.68

andals

-0.67

orks

-0.67

sung

-0.66

 toured

-0.64

 Volunte

-0.64

POSITIVE LOGITS

 posit

1.01

 inference

0.94

 deduction

0.90

 necessity

0.89

 fallacy

0.87

 deductions

0.86

 progression

0.86

istically

0.86

istic

0.84

\\\\\\\\

0.83

Activations Density 0.014%

adjectives and nouns relating to logical reasoning

logical reasoning and arguments

No Comments

No Known Activations