INDEX

Explanations

phrases related to causation or explanation

oai_token-act-pair · gpt-3.5-turbo

causal relationships and explanations

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

 bible

-0.67

 clipboard

-0.63

 wraps

-0.63

abases

-0.61

abytes

-0.60

letter

-0.59

 hoops

-0.59

 adjourn

-0.58

 mang

-0.57

 ages

-0.56

POSITIVE LOGITS

pez

0.67

riott

0.67

 Santos

0.67

WARD

0.65

 acute

0.64

endered

0.64

 Firstly

0.63

manuel

0.63

 anecd

0.62

owler

0.62

Activations Density 0.450%

phrases related to causation or explanation

causal relationships and explanations

No Comments

No Known Activations