INDEX

Explanations

legal or court-related terms

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

 livest

-0.84

pees

-0.81

pez

-0.73

resent

-0.72

emies

-0.72

emouth

-0.71

ower

-0.68

orks

-0.67

quire

-0.67

kered

-0.67

POSITIVE LOGITS

cases

0.92

 involving

0.87

esan

0.84

 Keen

0.77

BACK

0.74

McA

0.73

case

0.72

 cases

0.70

 scenario

0.68

cmp

0.66

Activations Density 1.524%

legal or court-related terms

No Comments

No Known Activations