INDEX

Explanations

words related to scientific and technical terms like "baseline," "experiment," and "equilibrium."

oai_token-act-pair · gpt-3.5-turbo

terms associated with baseline or equilibrium states

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

phan

-0.76

acles

-0.76

Else

-0.71

atom

-0.69

Office

-0.69

uild

-0.69

Anth

-0.68

paces

-0.67

office

-0.67

odes

-0.67

POSITIVE LOGITS

 baseline

0.98

phrine

0.83

 sidx

0.77

 jumper

0.74

 outline

0.71

 smoot

0.70

 dummy

0.70

 shif

0.67

 heartbeat

0.64

 level

0.63

Activations Density 0.009%

words related to scientific and technical terms like "baseline," "experiment," and "equilibrium."

terms associated with baseline or equilibrium states

No Comments

No Known Activations