INDEX

Explanations

terms related to team dynamics and team performance

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

hesis

-0.81

OGR

-0.79

chnology

-0.77

quez

-0.77

rift

-0.76

SS

-0.73

SPONSORED

-0.72

ombat

-0.71

à¥

-0.68

cue

-0.68

POSITIVE LOGITS

how

0.92

 halfway

0.84

 overcoming

0.73

 preserving

0.73

 reforming

0.73

 getting

0.73

 calib

0.72

 protecting

0.71

 improving

0.71

 whether

0.71

Activations Density 0.471%

terms related to team dynamics and team performance

No Comments

No Known Activations