INDEX

Explanations

verbs related to following or tracking

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

cer

-0.80

voy

-0.70

bodied

-0.69

cit

-0.69

atonin

-0.68

ãĥĨãĤ£

-0.67

pite

-0.66

venge

-0.66

cin

-0.65

anyon

-0.65

POSITIVE LOGITS

 suit

0.98

 closely

0.92

 directions

0.75

 footsteps

0.73

 instructions

0.72

Sym

0.70

SHIP

0.67

ansen

0.66

suit

0.65

 along

0.65

Activations Density 0.435%

verbs related to following or tracking

No Comments

No Known Activations